Nvidia's valuation is near the lows. BofA sees this as a growth opportunity

Bank of America analyst believes that Nvidia shares retain significant growth potential. According to him, the current valuation of the company has fallen to the levels from which the quotes used to bounce confidently. The pressure on the chipmaker increased fears that technology giants are developing their own AI-processors and can partially reduce demand for solutions Nvidia. On December 2, Nvidia securities rose in price for the second day in a row for the first time since mid-November.

Bank of America analyst Vivek Arya believes that shares of chipmaker Nvidia retain growth potential despite recent pressure due to concerns that major technology companies may switch to their own AI chips. Against this backdrop, Nvidia's securities in late November fell to their lowest level since September, MarketWatch writes.

BofA notes that Nvidia shares are now trading at a P/E (Price/Earnings) multiple of about 25. This ratio reflects the ratio of price to projected earnings for the next year and shows how expensive investors value the company's future growth. For Nvidia, the current level is close to the local lows seen in October 2023 and July 2022. However, previously, when the multiple fell to these lows, it usually recovered to 30-40 within three to six months, the bank analyst explains. Over the past five years, the median value has been 37. By comparison, the P/E ratio on the S&P 500 index is now around 24, meaning Nvidia's valuation is only slightly above the market average.

Investors in recent weeks have increasingly debated whether specialized solutions like Broadcom's ASIC chips could eventually reduce demand for general-purpose graphics processing units (GPUs) from Nvidia and AMD, MarketWatch explains. The occasion was the success and subsequent rise in Alphabet's stock: in November, Google unveiled its Gemini 3 AI model and emphasized that it was training it on its own TPU chips, which the company is developing with Broadcom. TPUs are just processors built for a specific task, they are optimized for certain types of AI computing and are used only inside Google's infrastructure.

An additional pressure factor was the record wide discount of Nvidia to Broadcom shares - about 40% against the historical norm of about 10%, notes MarketWatch. According to a BofA analyst, this difference reflects market expectations that Broadcom will be able to take away at least ten percentage points of Nvidia's AI market share in the second half of 2026 and in 2027.

All this has heightened the fears of Nvidia investors, but BofA believes that fears are exaggerated. First, the transition to alternative chips in the coming year is unlikely: the supply of high-performance processors remains limited, and Nvidia continues to receive priority capacity allocation from Taiwan Semiconductor Manufacturing Co. - Second, Google's TPUs have proven effective only inside the company's own data centers, while Nvidia's solutions are used much more widely and are suitable for a variety of AI models.

Morgan Stanley analyst Joseph Moore also believes that Nvidia will retain "dominant market share". At the same time, he admits: it is still unclear what factor will be able to reverse the negative investor sentiment, reports MarketWatch.

The Morgan Stanley team estimates that in 2026, Broadcom and AMD's AI chip revenue will grow slightly faster than Nvidia's - mainly because of supply chain specifics and the fact that Nvidia itself has already reached annual sales of about $205 billion. The shortage of key AI components, according to Moore, will continue at least through the end of next year.

In the next 12 months, says the analyst, the main worry for customers is to be able to get enough Nvidia products, especially with the release of the new Rubin platform - the next generation of AI chips.

"Of course the market needs alternatives, and in some tasks they will indeed be cost-effective," Moore says. - Google's TPU is by all accounts a worthy option that has already made notable contributions to a number of key models." However, the scale is not comparable: last quarter, Nvidia generated $51 billion in revenue in the data center segment - about 14 times more than Google's TPU revenue, the Morgan Stanley analyst notes.

In trading on December 2, Nvidia shares added about 2.7%. According to FactSet, if the positive dynamics continues, it will be the first two-day increase in quotes of the chipmaker in a row since November 10, Barron's points out .

They rose 1.7% on Monday following news that Nvidia plans to invest $2 billion in Synopsys, the largest developer of semiconductor design automation software. The move could expand the use of Nvidia's video chips in industrial and engineering applications.

In addition, technology company Hewlett Packard Enterprise announced the expansion of its partnership with Nvidia and the launch of an AI Factory Lab in France, where customers will be able to test their AI tasks on European infrastructure. Positive news also came from Japan: shares of industrial robot maker Fanuc jumped 6.5% after it was announced that the company will start adopting Nvidia's technology and jointly develop AI robots capable of executing voice commands.

This article was AI-translated and verified by a human editor